1. Field of the Invention
The present invention relates to a still video camera and, more particularly, to a still video camera including an automatic focus control apparatus.
2. Description of the Prior Art
A still video camera is a camera using a solid-state image pickup device (e.g., a CCD) as a light-receiving element to store image information as an electrical signal in an information recording medium such as a magnetic disk. In a conventional camera of this type, development of a film is not required since a silver chloride film is not used. In addition, image information can be transferred to a remote area. Therefore, the still video camera has received a great deal of attention in favor of a variety of image processing operations. Most of solid-state image pickup devices used in still video cameras in an initial stage of their development are primarily developed for VTR cameras. Mechanical shutters such as a focal plane shutter are required
When an object is photographed with an electronic flash by a still video camera of this type, an exposure amount must be controlled with high precision due to the following reason. When a solid-state image pickup device is a CCD, slight underexposure with respect to an optimal exposure amount causes excessive whitening of a bright portion of an image, and slight overexposure causes excessive blackening of its dark portion. In a conventional silver chloride film, even if an actual exposure amount is deviated from the optimal exposure amount, an error can be corrected during development and printing In a conventional still camera using a silver chloride film, a distance between the camera and the object is measured by auto focusing (automatic focusing control) in accordance with the following formula EQU Guide Number=Distance.times.F-number
The f-number is then calculated by the above formula, and exposure control can be relatively easily performed (so-called "flashmatic" control) In addition, the distance can be set in one of the steps between .infin. (infinite) and 1 m.
In a conventional still video camera using a CCD as a light-receiving element, optimal exposure control cannot be performed due to the narrow latitude range of the CCD. In the still video camera, exposure must be controlled with high precision. For example, a light control electronic flash is used to control the emission amount of the electronic flash.
FIG. 1 is a diagram showing an exposure control system of a conventional still video camera. When a trigger (emission start signal) is input to a light emission controller 1, the controller 1 operates an electronic flash 2. Upon operation of the electronic flash 2, an object 3 is illuminated with light from the electronic flash 2. Light reflected by the object 3 is incident on a light-receiving element 5 through a light-receiving lens 4. An integrator 6 integrates a photoelectric conversion output from the light-receiving element 5 simultaneously when the electronic flash 2 emits light. When an output from the integrator 6 reaches a light control level determined by CCD sensitivity and a selected f-number, a comparator 7 supplies a stop signal to the light emission controller 1. Therefore, the emission operation of the electronic flash 2 is stopped by the light emission controller 1.
FIG. 2 is a graph showing conversion characteristics of the electronic flash emission amount. Referring to FIG. 2, the electronic flash emission amount is plotted along the ordinate, and time t is plotted along the abscissa. A trigger is applied at time t1, and the electronic flash emission amount is abruptly increased, as is apparent from FIG. 2. When an integrated value of the integrator 6 reaches a light control level at time ts, the emission from the electronic flash 2 is stopped. An area indicated by hatched lines is an actual emission amount. A dotted curve is an emission curve of the electronic flash in a full emission state If the emission amount in the full emission state is zero at time t2, t2 can reach after ts in a duration between time t1 and time t2. A duration between time ts and time t2 is set to be a predetermined time interval (e.g., 1/60 second) in accordance with a given type of camera. The duration between time t1 and time t2 is a maximum integration time of the integrator 6.
A Xenon tube is used as an electronic flash. As shown in FIG. 2, when emission of the electronic flash 2 is interrupted during its emission period, the circuit arrangement of the light emission controller 1 is extremely complicated. A time error occurs between the generation of the emission stop signal and the end of actual emission For this reason, it is difficult to turn off the electronic flash with high precision during emission It is extremely difficult to turn off the electronic flash with high precision during the initial period of its operation. As a result, even if an automatic light control electronic flash is used in an electronic flash photographing mode wherein the distance is the closest focusing distance in a full-aperture state, the resultant image is degraded by underexposure and is often whitened In addition, a complicated circuit arrangement is required to result in a bulky system at high cost.
A solid-state image pickup device (frame-interline CCD or FIT-CCD) was recently developed. The FIT-CCD comprises a photosensor section, a vertical transfer section for vertically transferring an output from the photosensor section , and a memory for storing an output from the vertical transfer section FIG. 3 is a block diagram of the FIT-CCD. Referring to FIG. 3, reference numeral 10 denotes a photosensor section for converting an image signal into an electrical signal; 11, a vertical transfer section for vertically transferring the electrical signal from the photosensor section 10; and 12, a memory for storing an electrical signal (charge) transferred from the vertical transfer section 11. An output from the memory 12 is output through a horizontal transfer section 13.
The photosensor section 10 generates charge corresponding to input image information and supplies the charge to the vertical transfer section 11. The vertical transfer section 11 temporarily holds the charge supplied from the photosensor section 10 and transfers the held charge to the memory 12 in response to a shift clock. The memory 12 stores analog charge transferred from the vertical transfer section 11. The charge corresponding to the image information and transferred to the vertical transfer section 11 is no longer influenced by external light. The image information at the time of transfer is stored in the memory 12. In this sense, the FIT-CCD shown in FIG. 3 has an electronic shutter function An exposure amount of the FIT-CCD is controlled by changing timings of shift pulses for transferring the charge from the photosensor section 10 to the vertical transfer section 11. When such a characteristic of the FIT-CCD is properly utilized, light control during electronic flash operation of the still video camera can be performed.
The FIT-CCD was recently developed as a solid-state image pickup device, as described above. An electronic shutter function can be realized by using this FIT-CCD. The electronic shutter function will be described below.
FIG. 4 shows an arrangement of the FIT-CCD. Referring to FIG. 4, reference numerals 10a to 10d denote photosensor sections which are vertically aligned to generate changes in accordance with the amounts of received light. Each photosensor section is constituted by four photosensors. Sixteen photosensors L11 to L44 constituting the photosensor sections 10a to 10d are arranged to form a 4.times.4 matrix. Reference numerals 11a to 11d denote vertical transfer CCDs for receiving the charges from the corresponding photosensor sections and transferring the charges in the vertical direction Reference numeral 14 denotes a charge extraction drain for extracting the charges of the vertical transfer CCDs 11a to 11d. Reference numeral 15a denotes a first field storage section for storing the charge of the first field transferred from the vertical transfer CCDs 11a to 11d; and 15b, a second field storage section for storing the charge of the second field transferred from the vertical transfer CCDs 11a to 11d.
An image obtained from the photosensors L11 to L14 and L31 to L34 of the first and third rows (odd-numbered rows) is given as an image of the first field, and an image obtained from the photosensors L21 to L24 and L41 to L44 of the second and fourth rows (even-numbered rows) is given as an image of the second field The images of the first and second fields constitute a one-frame image.
Reference numeral 16 denotes a horizontal transfer CCD for horizontally transferring the charges stored in the first field storage section 15a and the second field storage section 15b and outputting them outside the storage sections 15a and 15b. Reference symbols .phi.V1 to .phi.V4 denote vertical register clocks; .phi.V5 to .phi.V8, storage register clocks for transferring the charges stored in the first and second field storage sections 15a and 15b to the horizontal transfer CCD 16; and .phi.H1 to .phi.H4, horizontal register clocks for horizontally outputting the charge from the horizontal transfer CCD 16. Reference symbol RS denotes a reset gate; OG, an output gate; RD, a reset drain; and OS, an output transistor source. Reference symbols OD1 and OD2 denote output transistor drains.
The operation of the FIT-CCD having the arrangement described above will be generally described below. The charges remaining in the photosensors L11 to L44 and the vertical transfer CCDs 11a to 11d are extracted to the charge extraction drain 14 and are initialized The initialized state corresponds to an open state of the electronic shutter. Light reflected by an object (not shown) is incident on the photosensors L11 to L44 of the photosensor sections 10a to 10d, and the photosensors L11 to L44 generate charges corresponding to the incident light beams. The charges generated by the photosensors L11 to L44 are transferred to the corresponding vertical transfer CCDs in response to shift clocks when a predetermined period of time has elapsed. When the charges are transferred from the photosensors L11 to L44 to the vertical transfer CCDs 11a to 11d, changes stored in the vertical transfer CCDs are no longer changed. This state corresponds to the closed state of the electronic shutter.
The charges stored in the vertical transfer CCDs 11a to 11d are shifted in response to shift clocks .phi.V1 to .phi.V4 and are stored in the first and second field storage sections 15a and 15b. More specifically, of charges stored in the photosensors L11 to L44, the charges of the photosensors of the first and third rows are stored in the first field storage section 15a, and the charges of the photosensors of the second and fourth rows are stored in the second field storage section 15b. The charges stored in the first and second field storage sections 15a and 15b are transferred to the horizontal transfer CCD 16 in response to the shift clocks .phi.V5 to .phi.V8. The charges are then output from the horizontal transfer CCD 16 to the outside of the CCD 16 in response to the shift clocks .phi.H1 and .phi.H2.
The frame photographing operation using the FIT-CCD will be described in detail with a timing chart of FIG. 5. Referring to FIG. 5, A represents the operation of the portion constituting the first field of the first and third rows, and B represents the operation of the portion constituting the second field of the second and fourth rows. Time is plotted along the abscissa.
Unnecessary charges of the vertical transfer CCDs 11a to 11d are extracted to the charge extraction drain 14 for a duration between time t0 and time t1. The unnecessary charges of the photosensors of the odd-numbered rows (i.e., photosensors (L1j and L3j where j=1 to 4) constituting the first field)) are transferred to the charge extraction drain 14 through the vertical transfer CCDs (the hatched region of A in FIG. 5) for a duration between time t1 and time t2. At the end of this charge transfer, i.e., at time t2, the exposure of the photosensors of the odd-numbered rows is started. In other words, time t2 is an electronic shutter open time. Similarly, the unnecessary charges of the photosensors of the even-numbered rows (i.e., the photosensors (L2j and L4j where j=1 to 4)) are transferred to the charge extraction drain 14 through the vertical transfer CCDs for a duration between time t2 and time t3 (the hatched region of B in FIG. 5). At the end of this charge transfer, i.e., at time t3, the exposure of the photosensors of the even-numbered rows is started. In other words, time t3 is an electronic shutter open time. The above operation corresponds to resetting of the photosensors.
When an exposure time t.sub.EXP =t4-t2=t5-t3 determined by brightness of the object has elapsed, the signal charges of the photosensors of the odd-numbered rows are instantaneously transferred to the vertical transfer CCDs 11a to 11d at time t4 (electronic shutter closing operation). The charges are transferred to the first field storage section 15a for a duration between time t4 and time t5. The signal changes of the even-numbered photosensors are instantaneously transferred to the vertical transfer CCDs 11a to 11d (electronic shutter closing operation). These charges are transferred to the second field storage section 15b for a duration between time t5 and time t6. Therefore, the signal charges of the odd-numbered photosensors are stored in the first field storage section 15a and the signal charges of the even-numbered photosensors are stored in the second field storage section 15b. The signal charges stored in the first and second field storage sections 15a and 15b are read out row by row by using the horizontal transfer CCD 16 as needed. The readout signals are recorded in, e.g., a 2" floppy disk.
When such an FIT-CCD is used as an image pickup element, exposure control during emission of the electronic flash can be performed. More specifically, when a predetermined period of time has elapsed after the electronic shutter is opened, emission of the electronic flash is started. The emission of the electronic flash is completed, light reflected by the object is received by the light-receiving element and is integrated. When the integrated value of the reflected light (i.e., an exposure amount) reaches a given value, the electronic shutter is closed while emission of the electronic flash continues.
As described above, in frame photographing in the electronic shutter operation using the FIT-CCD in normal exposure, the exposure timings are alternately different in adjacent rows. However, in the normal exposure mode, the exposure time t.sub.EXP is sufficiently long as compared with the charge transfer durations (i.e., t1 to t2 and t2 to t3). Exposure timing errors do not present serious problems.
However, electronic flash frame photographing using an FIT-CCD poses a problem. For example, when emission of the electronic flash is started at time t2, the radiation intervals of the electronic flash on the photosensors of the odd- and even-numbered rows differ from each other. Since the electronic flash emission time is a maximum of about 1 ms, the exposure timing error causes a difference between the exposure amounts. When the image is reproduced later, the difference appears as flickering. In other words, the image flickers.
FIG. 6 is a diagram showing a conventional still video camera using a CCD. Light from an object (not shown) is photoelectrically converted by a light-receiving element into an electrical signal. The electrical signal is amplified by an amplifier 21, and the amplified signal is supplied to a photometric circuit 22. The photometric circuit 22 measures a brightness level of the object in accordance with an output from the amplifier and supplies an object brightness signal to a system control circuit 23 The system control circuit 23 calculates an exposure amount and an f-number in accordance with the brightness of the object.
When the exposure amount and the f-number are calculated, the system control circuit 23 controls an aperture 25' to obtain an optimal aperture state At the same time, the system control circuit 23 controls an output timing of a timing generator 24 to control a shutter speed of a CCD camera unit 25. If the CCD camera unit 25 is constituted by an interline CCD, the time interval between the start of exposure of the light-receiving section and start of the transfer of the charge to the vertical transfer section is defined as the shutter time. If the shutter time is prolonged, the exposure amount is increased However, if the shutter time is shortened, the exposure amount is decreased.
An image signal received in the CCD camera unit 25 as described above is sent to a recording circuit 26. Color separation and luminance signal processing of the image signal are performed in the recording circuit 26. The processed image signal is then FM-modulated, and the FM-modulated signal is recorded in a magnetic disk 27. The magnetic disk 27 is rotated by a motor 29. The motor 29 is driven by a servo motor driver 28 under the control of the system control circuit 23. Therefore, the image is recorded in a predetermined area.
In the conventional camera as described above, the light-receiving element 20, the amplifier 21, and the photometric circuit 22 are required to measure the brightness of the object. Since the photometric light-receiving element and the like are required in the conventional camera, the camera becomes bulky and expensive. The photometric light-receiving element 20 must have the same characteristics as those of the CCD in the CCD camera unit 25. However, it is very difficult to obtain identical characteristics. For this reason, a photometric error occurs, and high-precision photometric operations cannot be performed.
Along with the unceasing advance of semiconductor techniques, various types of image processing apparatuses (e.g., a still camera and a video camera) using solid-state image pickup devices have been commercially available. In order to perform automatic exposure control in these apparatuses or automatically control exposure by driving an auto iris assembly, photometric information is inevitably required.
In a conventional still video camera, a photometric photosensor is arranged in addition to the solid-state image pickup device. An output from the photometric photosensor is used as photometric information.
In a conventional video camera, an output from the image pickup device is used. The outputs are integrated within a predetermined period of time, and the resultant voltage level is used as photometric information
In the still video camera, however, the additional or photometric photosensor complicates the arrangement. In addition, the mounting position of such a photosensor must be taken into consideration.
In an arrangement wherein the outputs from the image pickup device are integrated within a predetermined period of time and the voltage level of the integrated value is used as photometric information, this operation is effective when the image pickup device is continuously operated as in a video camera. However, in a still video camera, the image pickup device is intermittently driven, the above method is ineffective due to the following reason. When the auto iris assembly is driven to control the exposure, incident light controlled by the auto iris assembly is received by the image pickup device. Outputs from the image pickup device are integrated. If the integration level is not suitable, a correction value is fed back to the auto iris assembly to control the light amount again. It takes a long time to obtain a steady state. In addition, power consumption is undesirably increased.
It is possible to drive the image pickup device at high speed and its output serves as photometric information. In this case, an HCCD drive clock must have a frequency of about several tens MHz, thus requiring sophisticated circuit techniques. In addition, it is difficult to limit a photometric area, and power consumption is undesirably increased.
FIG. 7 is a diagram showing a conventional video camera using a solid-state image pickup device such as a CCD. Image information obtained by a CCD camera unit 25 is output to a signal processor 30 in response to timing pulses (transfer pulses) output from a timing generator 24. The signal processor 30 converts the input signal into a video signal of, e.g., an NTSC scheme.
The video signal output from the signal processor 30 can be monitored by an electric viewfinder (to be referred to as an EVF hereinafter) 31. When a switch SW (normally a release switch) is turned on to instruct the start of recording, the system controller 32 supplies control signals to the timing generator 24 and the signal processor 30 to cause them to perform the above-mentioned operations.
In the conventional video camera, images of two fields (one field is 1/60 second) constitute an image of one frame, i.e., 1/30 second, and a video signal representing a one-frame image is output, as shown in FIG. 8A. In other words, only the video mode can be set in the video camera. In order to obtain a still image in such a video camera, after the release button is depressed, as shown in FIG. 8B, a time lag is caused until the photographing signal shown in FIG. 8C rises. In other words, exposure is performed for a predetermined time interval after the release button is depressed. During this time interval, the shutter is opened and is then closed when an actual exposure amount reaches a predetermined exposure amount. Meanwhile, an image stored in the CCD is recorded.
In this case, the video output is not synchronized with the release button operation. Photographing must be started after an image is established in the video output, and a time lag is caused. If a still mode is set such that photographing is immediately started upon depression of the release button, a video signal is not output. A television display cannot be performed, and hence an EVF display cannot be undesirably performed.
FIG. 9 is a block diagram of a conventional still video camera using a CCD as an image pickup device. A CCD camera unit 25 outputs an image signal corresponding to an object in response to a timing pulse supplied from a timing generator 24 through a driver 33. Reference symbol .phi.V denotes a charge transfer pulse; and .phi.H, a signal read pulse. This image signal is input to and sampled and held by a sample/hold circuit 34. The sample/hold circuit 34 outputs a video signal. The output video signal is modulated by a recording circuit (not shown) The modulated signal is recorded in the magnetic disk.
An output from the sample/hold circuit 34 is also supplied to an A/D converter 35 and is converted into digital data. The digital data is stored in a semiconductor memory 37. In this case, since the access time of the semiconductor memory 37 is generally very long, the following implementation is required That is, the semiconductor memory 37 is divided into a plurality of blocks by a gate array 38. Read/write access of the image data is performed in units of blocks. More specifically, a latch section 36 is connected to the input of the semiconductor memory 37. The latch section 36 has latches corresponding to the number of memory blocks The outputs from the A/D converters 35 are sequentially latched in response to latch pulses from the gate array 38 and are written in the corresponding memory blocks.
Since the image data is output in synchronism with a read pulse .phi.H from the CCD camera unit 25, the image data must be written in the semiconductor memory 37 within the period of .phi.H. As shown in FIG. 8, in order to eliminate a difference between the H timing and the data write timing of the semiconductor memory 37, as shown in FIGS. 8B and 8C, timing control of the memory must be performed by the latch section 36 and the timing control gate array 38
FIG. 10 is a diagram showing the main part of such a still video camera. Referring to FIG. 10, reference numeral 40 denotes a recording circuit comprising an FM modulator 40a for FM-modulating the image signal (recording signal) from an image pickup means (not shown) and a recording amplifier 40b. Reference numeral 41 denotes a magnetic head assembly for recording a recording signal from the recording amplifier 40b in a magnetic sheet (e.g., a magnetic disk) 42 as a recording medium or reading out the recorded signal from the magnetic sheet 42. Reference numeral 43 denotes a servo control system comprising a head feed driver 43a for feeding the magnetic head assembly 41 to a predetermined position on the magnetic sheet 42, a driver 43b for a magnetic sheet rotation motor 44, and a servo controller 43c for controlling the drivers 43a and 43b. Reference numeral 45 denotes a detector (FG coil) for detecting rotation of the motor 44. Reference numeral 46 denotes a detector (PG coil) for detecting the center of the PG yoke in the magnetic sheet 42. Signals from the detectors 45 and 46 are supplied to the servo controller 43c. The servo controller 43c controls such that the speed of the magnetic sheet 42 is kept constant The servo controller 43c also controls a rotational phase of the magnetic sheet 42. Reference numeral 47 denotes a reproducing amplifier for amplifying an output read out by the magnetic head assembly 41; 48, an RF detector for RF-detecting an output from the reproducing amplifier 47; and 49, a system controller and clock generator.
FIG. 11 is a view for explaining a conventional recording circuit section in the still video camera. The magnetic head assembly 41 comprises a first head H1 and a second head H2. When frame recording is performed, the heads H1 and H2 are switched by a head switching circuit 50 such that field 1 is recorded by the head H1 in the magnetic sheet 42 and the head H2 records field 2 therein
FIG. 12 is a timing chart showing the recording signal from the image pickup means and the switching states of the heads H1 and H2.
In the conventional still video camera described above, as shown in FIG. 12, the head H1 is selected to record field 1, and the head H2 is selected to record field 2. In this manner, the two heads must be used to perform one-frame recording.
In the image pickup system (e.g., a still video camera and a video camera) using an image pickup tube or a solid-state image pickup device (e.g., CCD) as an image pickup means, the object must be photographed in an n-focus state Assume that a black-and-white fringe pattern shown in FIG. 13A is photographed. An output level of the image pickup means in the in-focus state is decreased in the black area but is increased in the white area. As shown in FIG. 13B, the black and white areas correspond to the lower and upper peaks of a sinusoidal wave, respectively However, when the object is in the out-of-focus state, the upper and lower peaks respectively corresponding to the white and black areas are not conspicuous, as shown in FIG. 13C.
Judging from the above result, when automatic focusing (auto focus) is performed, the lens position is determined such that the amplitude of the high-frequency component is maximum, as shown in FIG. 13B. One of the automatic focusing techniques is a so-called "climbing" technique.
FIG. 14 is a view for explaining the operation according to the "climbing" technique. Referring to FIG. 14, the integrated values of the high-frequency components throughout one frame are plotted along the ordinate, and the distances from .infin.(infinite) to the closest focusing distance are plotted along the abscissa. According to the "climbing" technique, the lens is driven from a position near the .infin.position to a given point, e.g , L.sub.i. At the point L.sub.i, the lens is driven by a small distance .DELTA.lin the forward and reverse directions, and a change in integrated value of the high-frequency component is detected. If the integrated value of the high-frequency component at the point L.sub.i is larger than that at a point near the infinite position, and if the integrated value of the high-frequency component at the point L.sub.i is smaller than that at a point near the closest focusing point, the peak of the integrated value is located closer to the closest focusing distance point with respect to the point L.sub.i. The above operations are repeated at a point L.sub.i+1 spaced apart from the point L.sub.i by a predetermined distance. The repetition of the above operations allows detection of a point L.sub.f at which the integrated value is the largest The point L.sub.f is defined as an in-focus point. Note that the scanning direction may be a direction from the closest focusing distance to the infinite. The above "climbing" technique is used in a video camera or the like.
Another automatic focusing control technique utilizes the principle of trigonometrical survey in which the lens is scanned from the closest focusing distance to the infinite (or vice versa) to obtain a phase difference.
In a system such as a video camera in which video signals are constantly output, if the object distance is changed, the lens is moved to always obtain the in-focus state. Therefore, a technique for scanning the lens from the closest focusing distance to the infinite distance to obtain an in-focus position cannot be employed. Therefore, the "climbing" technique must be employed instead. The lens is slightly moved in the forward and reverse directions and the peak is detected in accordance with the lens driving direction and the change in high-frequency component. For this reason, if two peaks, i.e., high and low peaks as shown in FIG. 15 are present, the low peak (L.sub.1) is erroneously detected as the in-focus point although the high peak (L.sub.2) is present. In the phase comparison technique utilizing the principle of triangulation, a movable member such as a movable mirror in addition to the lens is required to result in an expensive, complicated system.
The "climbing" technique in which the peak of the high-frequency component of the video signal is detected to determine the in-focus point is a good auto focus technique except for some objects (e.g., a wall) without contrast. According to the "climbing" technique, when the lens is scanned from the closest focusing distance to the infinite distance, the lens can be accurately set in the in-focus position. In the still video camera, the in-focus state need not be always obtained unlike the video camera. In other words, the in-focus state must be obtained in the still video camera during only photographing. In addition, since the lens is moved for focusing, an additional mechanism is not required for scanning.